Spintronic devices are promising alternatives for next-generation computation accelerators, particularly in Ising machines, due to their potential for parallelization and low power consumption. These devices can operate at room temperature with low operational costs, making them suitable for solving complex optimization problems. The paper explores various spintronic-based Ising machines, including those based on spintronic oscillators, probabilistic bits, and spin-wave Ising machines. Spintronic oscillators, such as spin-torque nano-oscillators (STNOs) and spin-Hall nano-oscillators (SHNOs), are used to create artificial spin states through phase binarization. Probabilistic Ising machines (pIMs) leverage the stochastic nature of spintronic devices to solve optimization problems by sampling from probability distributions. Spin-wave Ising machines (SWIMs) utilize spin waves for efficient computation, offering advantages in scalability and thermal stability. The paper discusses the challenges and opportunities of different physical platforms, emphasizing the importance of control mechanisms and coupling strategies. It also highlights the potential of spintronic devices in overcoming the limitations of conventional computing, such as energy consumption and sequential processing. The study concludes that spintronic devices offer significant advantages in terms of miniaturization, operational frequency, and power efficiency, making them strong candidates for future computation accelerators. The integration of spintronic devices with machine learning and other emerging technologies is also discussed, highlighting their potential for real-world applications in complex optimization tasks.Spintronic devices are promising alternatives for next-generation computation accelerators, particularly in Ising machines, due to their potential for parallelization and low power consumption. These devices can operate at room temperature with low operational costs, making them suitable for solving complex optimization problems. The paper explores various spintronic-based Ising machines, including those based on spintronic oscillators, probabilistic bits, and spin-wave Ising machines. Spintronic oscillators, such as spin-torque nano-oscillators (STNOs) and spin-Hall nano-oscillators (SHNOs), are used to create artificial spin states through phase binarization. Probabilistic Ising machines (pIMs) leverage the stochastic nature of spintronic devices to solve optimization problems by sampling from probability distributions. Spin-wave Ising machines (SWIMs) utilize spin waves for efficient computation, offering advantages in scalability and thermal stability. The paper discusses the challenges and opportunities of different physical platforms, emphasizing the importance of control mechanisms and coupling strategies. It also highlights the potential of spintronic devices in overcoming the limitations of conventional computing, such as energy consumption and sequential processing. The study concludes that spintronic devices offer significant advantages in terms of miniaturization, operational frequency, and power efficiency, making them strong candidates for future computation accelerators. The integration of spintronic devices with machine learning and other emerging technologies is also discussed, highlighting their potential for real-world applications in complex optimization tasks.